The United Arab Emirates (UAE) and France this week announced plans for a one-gigawatt AI datacenter campus dedicated to advancing development of artificial intelligence.
The datacenter project, which will be built in France, was unveiled Thursday evening during a meeting between French President Emmanuel Macron and UAE President Mohamed bin Zayed Al Nahyan.
At one gigawatt in capacity, the facility will be enormous, dwarfing more traditional cloud and hyperscale campuses, which typically top out in the tens of megawatts. With that said, facilities on this scale aren’t unheard of, given the insatiable appetite of AI workloads for power-hungry GPUs and high-end networking kits. In the US, social media giant Meta recently began construction of a 2.3-gigawatt facility in Richland Parish, Louisiana.
Funding for the Franco-Emirati project will be sourced from a number of backers, reportedly topping €30-€50 billion ($31-$51 billion) — a fraction of the eye-watering $500 billion figure associated with the United States’ Stargate project — if that ever materializes. MGX, a $100 billion Abu Dhabi investment fund, is reportedly playing a role in financing both the French datacenter project and the ambitious Stargate initiative.
The first phase of UAE investments will be announced during the Choose France 2025 summit later this year, the nations said in a joint statement.
Along with supporting the expansion of datacenter infrastructure, the partnership will also see the UAE and France work to acquire state-of-the-art semiconductors, talent, and establish virtual data embassies for the development of sovereign AI and cloud services in both countries.
The tie-up isn’t surprising. The UAE firms like G42 have courted several high profile AI infrastructure providers including Microsoft and Cerebras among others, as the country has sought to insulate itself from increasingly restrictive export controls on American-designed accelerators crucial to AI development. The latest export restrictions set forth by the Biden administration would severely limit the number of accelerators available to the UAE, and much of the world, if they’re implemented by the Trump administration.
France, for its part, is home to one of Europe’s most influential model devs Mistral.AI. The startup has managed to produce large language models — the kind used to power AI chatbots like ChatGPT or Gemini — that are competitive with its larger, better funded rivals in the US.
SoftBank woos OpenAI with $40B, making Microsoft’s $13B look quaint
Datacenter energy use to more than double by 2030 thanks to AI’s insatiable thirst
Hardware quality problems and server supply chain kinks slow Amazon’s $100 billion AI build
Who’s afraid of DeepSeek’s impact on AI hardware sales? Not AMD CEO Lisa Su
Speaking of Mistral, alongside the Franco-Emirati partnership, Cerebras, which is backed by UAE AI darling G42, revealed it’s providing its wafer-scale compute platform to the French model builder. Available on Mistral’s Le Chat platform, “Flash Answers” promises generation speeds of 1,100 tokens a second, though we’ll note that Cerebra is juicing the numbers here a bit using speculative decoding.
These announcements come on the eve of the Artificial Intelligence Action Summit in Paris, which kicks off next week, where recent developments surrounding Chinese model builder DeepSeek, and newly elected US president Donald Trump are expected to be key points of discussion. ®
GIPHY App Key not set. Please check settings